There is a question almost nobody asks before producing a course, and it is the one that determines everything else: what will this person be able to do on Monday that they can’t do today? Answering it precisely — and then building the path to get there — is what instructional design is. It’s not the pretty part of the process (that’s production), but it is the part that decides whether the money you’re about to invest turns into real change or into one more file sitting on the platform.
Instructional design is the craft of planning how something will be taught so that the learner ends up able to do it. It defines what gets taught and what gets left out, in what order, with how much practice, with which examples, in what format, and how you verify that learning actually happened. It is to a course what a blueprint is to a building: you can build without one — plenty of people do — and the result gives it away.
It helps to think of it by contrast. When a company decides to train, it usually jumps straight to production: gather the expert, record what they know, build a slide deck, upload it to the platform. The content is there — sometimes excellent content — but nobody decided what the participant should walk away able to do, how much they needed to practice, or how anyone would know it worked. Instructional design is precisely all of that prior thinking, and it is invisible in the final product: when it’s done well, the course simply feels clear.
The professional behind it is called an instructional designer, and the role is still rare inside most companies — which is why the work tends to land on whoever has time, and that is almost never whoever has the craft. It’s a profession with decades of research behind it on how adults learn, how much information working memory can hold, and why spaced practice beats cramming.
Four concrete things, all visible to anyone taking the course even if they can’t name them:
An expert wants to tell you everything, because to them everything is interesting. Design decides what stays based on the objective, and that pruning usually cuts the course in half — which also lowers production costs and dropout.
Instead of multiple-choice questions about definitions, learners face the decisions they will actually make in their role — with the same incomplete data they’ll have when they make them for real.
Every screen leaves a question open that the next one answers. That tension — borrowed from the storyteller’s craft — is what holds attention when nobody is watching.
It’s designed before the content, not after, and it measures performance: if the person can make the right call on a case, it doesn’t matter whether they remember the textbook definition.
This is the path every course we produce follows. Each step ends in something you can read and correct, because changing an idea in a document costs minutes and changing it in a finished course costs weeks.
There are four, and all of them stay in your hands when the project ends. We insist on this because it’s the difference between buying a course and building a capability: with these documents, your team can maintain and expand the content without depending on us.
The course’s executive summary: objectives, audience, duration, module structure and how it’s assessed. It’s the document you share with leadership and with auditors.
The complete course in writing, screen by screen, with interactions and feedback. It’s the blueprint the course is produced from and the document it gets updated with later.
Cases, rubrics and passing criteria, with the logic of why each answer is correct — which is the part that turns an exam into a learning tool.
The format and the criteria so your team can design the next course following the same logic. It’s what lets the academy grow without starting from scratch.
They aren’t our inventions: they come from decades of research on adult learning. What is ours is the stubbornness with which we defend them when deadlines squeeze.
All of the above sounds reasonable on a page; the fair thing is for you to check. In our portfolio you can open real lessons — one on a subject as dry as calculating insurance compensation, another on AI-powered recruiting — and see how these principles behave once they stop being theory.
“Our academy has been a success, both financially and educationally, thanks to their remarkable pedagogical work.”
Behind every design is Oliver Barona, a senior instructional designer with more than fifteen years in the craft and winner of the 2025 iSpring International Course Creation Contest. The full method is explained on the method page.
When the cost of people not learning is higher than the cost of designing well — which happens more often than it seems. If twenty people will take the course once, gathering the expert and recording them may be enough; but if every new hire will take it for the next three years, if the mistake it prevents costs money, or if the subject is hard enough that people abandon it halfway through, then the design pays for itself.
There is another signal, less obvious and more common: when the course was already attempted internally and stalled halfway. It’s almost never a lack of content or of will — what’s missing is someone to decide what gets left out, and that decision is uncomfortable to make from the inside.
It’s the craft of deciding how something will be taught so the person can do it afterwards: what gets taught, in what order, with how much practice, with which examples, and how you verify they learned. It’s the equivalent of the architect on a construction site — someone can lay bricks without blueprints, but the result shows. In a course, instructional design is what separates an experience that changes Monday’s work from a slide deck with an exam at the end.
Your expert knows how to do the job, and that’s exactly what makes it hard for them to teach it: when you master something, you forget how much it cost you to learn and treat as obvious twenty steps that aren’t obvious to someone new. Instructional design extracts that knowledge, orders it for someone who doesn’t have it, decides what gets practiced and what gets looked up, and turns experience into a sequence another person can walk through on their own.
It’s the blueprint of a course before it’s produced: screen by screen, what’s said, what’s shown, what the participant does and what happens when they get it wrong. It’s useful for two very practical things: it lets you review and correct the course while changing something costs minutes — instead of discovering the problem once everything is recorded and programmed — and it gives your team a document they can use to update the content in the future without hiring us again.
ADDIE is the classic model of the craft, and its letters describe five phases: analysis, design, development, implementation and evaluation. We do use it as a backbone, because the order it proposes is sensible, but not in its rigid textbook version: we work in short iterations, showing navigable pieces early instead of saving everything for a grand reveal at the end. Models are for thinking, not for justifying six months of silence.
In terms of what the person will be able to do, not what they will know. “Know the credit policy” is not an objective, because nobody can verify it; “decide whether a credit application meets the criteria and justify the decision” is, because it can be observed and assessed. That change in wording looks cosmetic and isn’t: by forcing you to name an observable action, it automatically defines what must be practiced and how it will be assessed.
We use it in production — to generate case variants, draft first versions and speed up materials — but the design decisions are made by a person: what gets taught, what gets practiced, what gets assessed and what gets cut. The reason is simple: those decisions depend on knowing your operation and where it hurts, and that comes out of the diagnostic and the conversations with your people, not out of a model.
Tell us what you need to teach and who knows how to do it today. We’ll tell you how we would design it, which deliverables you’d keep, and in what order.